Does Nursing Behaviour of Sows in Loose-Housing Pens Differ from That of Sows in Farrowing Pens with Crates?

SIMPLE SUMMARY: Around birth and during lactation, sows are often confined to narrow farrowing crates in order to restrict their movements to prevent piglet crushing by sows. Since these housing systems are increasingly questioned for sow welfare reasons, more research is being carried out on alternatives such as loose-housing farrowing systems. As one central aspect of sows’ maternal behaviour, nursing behaviour impacts the mortality rate and growth performance of the piglets. Since maternal behaviour is, in turn, affected by the housing environment, this study aimed to compare a farrowing pen without sow confinement to one with a crate regarding the nursing behaviour of sows. The study results underline that sows’ nursing behaviour was influenced by the housing system and, furthermore, that the nursing behaviour of loose-housed sows was more similar to that of sows in semi-natural conditions. Unconfined sows terminated more nursing bouts and nursed piglets for a shorter period. The nursing frequency decreased in the course of lactation in both systems, whereas the proportion of sow-terminated nursing bouts increased in the four-week examination period. These might be the first steps in the natural weaning of piglets. ABSTRACT: Sows confined to farrowing crates are restricted in performing natural behaviour such as maternal behaviour. Loose-housing farrowing pens (LH) and farrowing pens with crates (FC) were compared regarding sows’ nursing behaviour via video analyses over four weeks per batch (one day per week). Nursing frequency was similar in LH and FC pens (1.25 ± 0.82 vs. 1.19 ± 0.75 nursings/sow/hour; p > 0.05). However, nursing duration differed between the two systems (LH: 5.7 ± 4.6 min vs. FC: 7.0 ± 5.0 min; odds ratio (OR) 1.168, p = 0.011). In LH pens, more nursing bouts were sow-terminated than in FC pens (OR 0.427, p = 0.001). The probability of sow-terminated nursing occurring increased from week 1 to week 4 (OR 3.479, adjusted p (p(adj)) < 0.001), while that of observing unnursed piglets decreased from week 1 to week 4 (OR 0.301, p(adj) < 0.001) and rose with increasing litter size (OR 1.174, p = 0.010). We conclude that nursing behaviour was affected by the farrowing system, with shorter nursing duration and more nursing terminations by the sow in LH than in FC pens. Since this corresponds to the nursing behaviour of sows in semi-natural conditions, it can be assumed that sows in LH pens are more likely to exhibit natural nursing behaviour.